 |
||||
|
||||
|
|
|
 |
||
|
||||
      |
GLERL 2000 Milestone Reports
GOAL: SUSTAIN HEALTHY COASTS
OBJECTIVE 2: Promote Clean Coastal Waters to Sustain Living Marine resources and to Ensure Safe Recreation, Healthy Seafood, and Economic Vitality
PM: Develop better tools, predictive models, and understanding related to water quality and coastal ecosystems management.
Milestone: Develop new perspective on the wind wave generation and growth processes.
Scientist: Paul C. Liu
Purpose: This work is a collaborative effort between GLERL (Paul C. Liu and David J. Schwab) and the Coastal Hydraulics Laboratory of U.S. Army Corps of Engineers (R. E. Jensen). The objective is to elucidate diversified perspectives on the wind wave generation and growth processes through application, comparison and assessment of the various presently available wind wave models.
Efforts: We conducted our assessment of the present status of the wind wave modeling by applying several frequently used wind wave models to a case of active wave generation and growth in Lake Michigan and comparing the results with actual measurements. One of the models used is the widely used, state-of-the-art WAM model that generally requires supercomputer operation. Another, by contrast, is a simple parametric GLERL model that can be implemented on any basic computer system. Also included are two other models developed by the U.S. Army Corps of Engineers that are basically forerunners to the WAM model. All of the models are based on the concept of a wind-wave energy spectrum, which grows and decays in response to changes in the wind field. Our effort here is not on model inter-comparison, or model validation, but rather to address the question whether wind wave models based on the concept of a wave energy spectrum have reached their limits in terms of simulating the observed natural characteristics of wind waves. We start by presenting the details of the several models, applying them to a ten-day episode of measured wind and wave fields in Lake Michigan, and proceed to discuss the merits and shortcomings of the current state of wind wave modeling. The results led us to speculate that the assumption of wind wave spectrum may be the limiting factor in the further development toward more advanced numerical wave prediction models with significantly improved accuracy.
Customers:Atmospheric and oceanic scientists and engineers as well as marine wind and wave forecasters who are objectively interested in the state of wind wave forecasting and on the need for enhancement and improvements of current wind wave models will use the information provided.
Significance: The results of our analysis revealed a rather surprising and not previously recognized fact. The fact that the four disparate wind wave models produce results that are actually more similar to each other than they are to the measured data. This is particularly interesting considering the dissimilarity in physical contents between the WAM model, which is known to embody the most complex physics, and the GLERL model, which includes only a single empirical wind stress term. The results can be illustrated in the following two figures of correlation coefficient comparisons. It is readily observed that the correlation coefficients between the models are all higher than between model and measurements. An immediate plausible implication would be the present model developments have reached their limitation, further improvement in the present models will be unlikely from further refinement or validation. To a large extent, fresh and totally new approaches to wave modeling will be required for further substantial improvement.
Figure 1: Mean correlation coefficients for significant wave height between model and measurement (the red bars) and between model and model (the yellow bars). The correlation coefficients between model and model are all higher than that between model and measurement. (The four models used are WAM model developed by the international WAveModeling group, SHALW for SHALlow-water Wave model and DWAVE for Deep water WAVE model are developed by Coastal Hydraulics Laboratory, U.S. Army Corps of Engineers, GLERL is the GLERL wave model developed by Mark Donelan at Canada Water of Inland Waters and modified and made operational at GLERL.) The model versus model comparisons evaluated the model under study and shows the average comparison with the three other models for each case.
Figure 2: Mean correlation coefficients for dominant wave period between model and measurement (the red bars) and between model and model (the yellow bars). The correlation coefficients between model and model are, again, all higher than that between model and measurement.
Success: The task of this collaboration effort has been completed. A manuscript has been prepared, submitted to the International Journal of Ocean Engineering, and accepted for publication.
Next Steps: At several points in time during the 20th century, the wind waves problem has been considered as solved. At several stages, however, the same problem had been considered as far from understood. At the threshold of a new century, both viewpoints are still coexist. This completed task led us to conclude that there is an urgent need for a totally fresh start with systematic and detailed wind and wave measurement toward thoroughly understanding of physics and processes in order to develop totally new, fresh, unpretended, and successful approaches for truly understanding of wind wave and modeling.
References
Liu, P.C., D.J. Schwab, and R.E. Jensen (2001): Has wind wave modeling reached its limit? (In press, Ocean Engineering.)
Last updated: July 9, 2002 mbl